The long-term goal of these investigations is to understand how genes cause congenital heart disease (CHD). The first set of objectives involved devising methods of analysis by unifying two approaches: the focus of the current application, involves obtaining clinical data and extending the theory to the analysis of these data. These sets of objectives are clearly synergistic, and examination of empiric data will continue to stimulate fresh ideas for theory, just as exploration of models has pointed to potentially revealing avenues of clinical study. Considerable progress in extending and developing the methods of genetic analysis for CHD has been achieved. Topics sufficiently explored to warrant extension to empiric data include: epistasis; selection; multicompartmental models; bingo models; threshold and tetrachoric models; circular distributions; and, angular homeostasis. The following disorders will be studied. (1) Cardiovascular malrotation in its several forms will illustrate aspects of angular homeostasis; theory predicts how extracardiac manifestations might be associated with some, but not all, types of maldevelopment of malrotation. (2) Papillary muscle position on the ventricular endocardium can be studied by directional statistics applied to echocardiographic and magnetic resonance images. Malposition has been related to flow lesions such as aortic stenosis. Papillary muscle position will be determined by echocardiography, and genetic factors studied in normal families, and in families ascertained through occurrence of CHD flow lesions, especially aortic stenosis. (3) The mild end of the flow lesion spectrum includes bicuspid aortic valve and aortic coarctation; often associated with these lesions is dilatation of the proximal aorta. The possibility that predisposition to these lesions is an autosomal dominant trait will be studied by searching, by magnetic resonance imaging and echocardiography, for subclinical anatomic alterations in relatives of probands. Development of extensive pedigrees will enable investigation of epistasis. (4) The occurrence and progression of aortic root dilatation in the Marfan syndrome, hereditary annuloaortic ectasia, and congenital bicuspid aortic valve will be studied as multiple-hit systems by means of bingo-gamma models. The genetic and evolutionary fitness of the Marfan phenotype can be explored by the method of branching processes. (5) Mitral valve prolapse is thought to be an autosomal dominant trait in some families. Probands will be categorized as to the presence of redundancy of valve apparatus, and relatives studied by echocardiography, with the dual aims of examining mendelian models and exploring ascertainment bias in an age-dependent trait. (6) The conjunction of digital abnormalities and CHD in certain mendelian conditions, notably the Ellis-van Creveld and the Holt-Oram syndromes, provides a way to test pathogenetic models based on quadrangular correction functions.